Completed
New ComfyUI installer CUDA 13, Torch 2.9.1, Triton + attention libs
Class Central Classrooms beta
YouTube videos curated by Class Central.
Classroom Contents
NVFP4 with CUDA 13 Full Tutorial - 100%+ Speed Gain, Quality Comparison and New Cheap Cloud SimplePod
Automatically move to the next video in the Classroom when playback concludes
- 1 New ComfyUI installer CUDA 13, Torch 2.9.1, Triton + attention libs
- 2 NVFP4 speedup claims vs real tests; why CUDA 13 enables new models
- 3 Prebuilt FlashAttention/SageAttention/xFormers for many GPUs Windows + Linux
- 4 Quality roadmap: FLUX2 Dev, Z Image Turbo, FLUX Dev BF16/FP8/GGUF/NVFP4
- 5 Downloader adds NVFP4: FLUX2 Dev, FLUX Dev Context/Dev, Z Image Turbo
- 6 SimplePod AI intro: RunPod-style pods, cheaper rates, permanent storage
- 7 Musubi Tuner FP8 Scaled: quality myths vs GGUF + why scaled matters
- 8 Quantization & precision FP32/BF16/FP8/GGUF + Qwen3 low-VRAM encoders
- 9 ComfyUI v73 zip: CUDA 13 included; update NVIDIA drivers only v72 deprecated
- 10 Update steps: overwrite zip, delete venv, run install/update .bat
- 11 Python: 3.10 recommended supports 3.10-3.13; fresh vs update
- 12 New installer flow: uv speed, standalone use, backend libs detected
- 13 Stability flags: --cache-none vs --disable-smart-memory OOM/stuck fixes
- 14 SwarmUI presets: 32 presets supported; drag/drop + auto model downloader
- 15 Update SwarmUI model-downloader zip extract + overwrite
- 16 Download bundles/models Z Image Turbo Core + NVFP4 options
- 17 Update/launch SwarmUI; point to updated ComfyUI backend + set args
- 18 Live gen test: Z Image Turbo BF16 @1536x1536
- 19 Switch to NVFP4: VRAM cache behavior; 1024x1024
- 20 FLUX2 Dev quality: FP8 Scaled vs NVFP4 side-by-side comparisons
- 21 Speed chart: FLUX2 NVFP4 about 193% faster than FP8 Scaled
- 22 Z Image Turbo quality: BF16 vs NVFP4 vs FP8 Scaled quant method
- 23 FLUX Dev: FP8 Scaled approx GGUF Q8; NVFP4 currently shows degradation
- 24 What precision means + model size examples FP32/BF16/FP8 Scaled/NVFP4
- 25 Practical recommendations: BF16 best; avoid FP16; raw FP8 vs FP8 Scaled
- 26 GGUF explained: block quant, slower runtime; use only when RAM is too low
- 27 Precision hierarchy recap + when to pick FP8 mixed/scaled over GGUF
- 28 SimplePod setup: register, add credits, open template link
- 29 Template config + RunPod price comparison disk, ports, GPU selection
- 30 Persistent volume: create + mount to /workspace
- 31 Launch RTX Pro 6000 pod; SimplePod vs RunPod pricing differences
- 32 Temp vs persistent disk: deleting instance wipes temp data - backup!
- 33 JupyterLab: upload zips, apt install zip, unzip ComfyUI in workspace
- 34 Run install script; unzip SwarmUI; start the model downloader
- 35 Downloader path for ComfyUI + folder structure; download Z Image Turbo bundle
- 36 Start ComfyUI; confirm CUDA 13 + Torch 2.9.1; connect via port 3000 Direct
- 37 Preset demo: Z Image Turbo Quality 1; fix VAE path; monitor VRAM
- 38 File Browser Direct: download outputs/models fast; upload files back
- 39 Restart server; install/start SwarmUI; open Cloudflared URL
- 40 SwarmUI backend: /workspace/ComfyUI/main.py + args; import presets
- 41 Download FLUX2 Core + NVFP4; share model paths between SwarmUI & ComfyUI
- 42 FLUX2 NVFP4 generation @2048x2048; VRAM usage + step speed
- 43 Cloud GPU pitfall: diagnosing a power-capped GPU
- 44 Resume: re-run template w/ volume; reconnect fast
- 45 Wrap-up: SimplePod pros direct/secure, cheaper storage
